Method, calibration and measurement device for determining a measured value

Abstract

To reduce measurement errors, the setup includes an input sensor (S1) for measuring an input quantity (u1) of a technical system (2) and for outputting an input sensor response (SA1), an input conversion unit (W1) for converting the input sensor response (SA1) into an input measurement value (y1), an output sensor (S2) for measuring an output quantity (u2) of the technical system (2) and for outputting an output sensor response (SA2), and an output conversion unit (W2) for converting the output sensor response (SA2) into an output measurement value (y2). During a calibration period in which the technical system (2) is in a calibration state, an input measurement value (y1) and an output measurement value (y2) are acquired and compared to generate a calibration reference value (ey).The input conversion unit (W1) is corrected to a corrected input conversion unit (W1) depending on the calibration reference value (ey), and in a temporal active section in which the technical system (2) is in an active state different from the calibration state, the corrected input conversion unit (W1) is used to measure the input quantity (u1).

Description

[0001] The present invention relates to a method for determining a measured value, wherein an input sensor for metrologically acquiring an input variable of a technical system and for outputting an input sensor response as a function of the input variable, an input conversion unit for converting the input sensor response into an input measured value, an output sensor for metrologically acquiring an output variable of the technical system and for outputting an output sensor response as a function of the output variable, and an output conversion unit for converting the output sensor response into an output measured value are provided. The invention further relates to a calibration device and a measuring device.

[0002] Power converter assemblies are known in a wide variety of designs and are used in a multitude of different applications. For example, electric and hybrid vehicle drives incorporate power converter assemblies to convert the direct current (DC) voltages supplied by batteries into suitable alternating current (AC) voltages. These AC voltages are then applied to the electrical machines to be driven. Power converter assemblies typically comprise combinations of rectifiers (AC / DC converters), inverters (DC / AC converters), and / or converters (DC / DC converters, AC / AC converters, boost converters, buck converters).

[0003] To test modern drive concepts that are at least partially based on electric motors, series topologies consisting of AC / DC converters, DC / DC converters, and DC / AC converters are typically used. A AC / DC converter first generates a DC link voltage from a generally multi-phase AC supply voltage (mains AC voltage). This DC link voltage can then be adjusted to a suitable voltage level by a downstream DC / DC converter (boost converter, buck converter). When using active and therefore switchable AC / DC converters ("synchronous rectifiers"), a separate downstream DC / DC converter can be omitted, and a regulated and thus appropriately adapted DC link voltage can be provided directly by the AC / DC converter.A DC voltage generated in this way, with an adjusted voltage level, can be converted into a suitable AC voltage using an inverter if AC voltages are required for testing, for example, an AC voltage for each phase of an AC electric motor under test. The DC or AC voltages obtained in this way are then applied, for example, to the stator windings of an electric motor under test to generate the necessary phase current for producing torque and / or speed. Corresponding structures are well known from the prior art, for example, from AT 523580 B1 or EP 2 855 193 B1.

[0004] In the aforementioned applications with serial topologies of rectifiers and inverters, it is necessary to precisely regulate the DC link voltage to a predefined setpoint to ensure satisfactory operation of a downstream inverter that further processes the DC link voltage. For this purpose, highly accurate and high-resolution sensors are typically used to measure the DC link voltage. However, in the case of using active rectifiers to convert mains AC voltage into a regulated DC link voltage, the above statements also apply to measuring the mains AC voltage applied to the rectifier on the mains side.If a mains AC voltage is measured incorrectly, and a controller for regulating the DC link voltage—and thus for switching the semiconductor switches of the mains rectifier—receives incorrect information about the mains AC voltage, the control performance is obviously impaired, accuracy is lost, and sometimes even stability problems occur. This is referred to as a reduced quality of the "AC-side mains control."

[0005] In the scenario described above, the calibration and adjustment of AC measurement chains regularly pose problems. It is well known that AC sensors and their corresponding measurement chains must be checked and, if necessary, calibrated and / or adjusted at regular intervals, for example, during commissioning to compensate for component variations, or during operation to compensate for aging effects and / or component drift. Calibration and correction of DC measurement chains are typically possible with minimal effort and without special technical equipment, allowing DC measurement chains for the intermediate circuit voltage to be precisely adjusted, e.g., during operational breaks or even during operation, to meet the required accuracy standards. However, this is not the case for AC measurement chains used to measure mains-side AC voltages.In fact, the calibration process for AC quantities is significantly more complex than for DC quantities. This is due, on the one hand, to the more complex time-dependent behavior of AC quantities (high voltages result in high signal gradients and high signal amplitudes), which necessitates more complex measurement chains for signal processing. On the other hand, particularly in cases involving high electrical power, it is difficult to generate suitable reference signals with known signal parameters that can be compared with measurement results from AC sensors. Consequently, situations frequently arise where the AC mains voltages applied as inputs to mains rectifiers are measured incorrectly or only inadequately.

[0006] In other technical applications with other technical systems where input and output quantities must be measured precisely, e.g., in test benches, (electric) motors, or energy converters in general, etc., undesirable measurement errors frequently occur in connection with rapidly changing alternating quantities. In particular, when alternating quantities are present on the input side of a technical system and direct quantities on the output side, comparable problems arise during calibration and measurement, almost regardless of the specific design of the technical system.

[0007] Therefore, one of the purposes of the present invention is to reduce measurement errors.

[0008] According to the invention, this problem is solved for the aforementioned method and the aforementioned calibration and measuring device by acquiring and comparing an input measured value and an output measured value during a calibration period in which the technical system is in a calibration state, preferably by subtracting them from each other or comparing them in another way to generate a calibration reference value, by correcting the input conversion unit to a corrected input conversion unit depending on the calibration reference value and / or by correcting the output conversion unit to a corrected output conversion unit depending on the calibration reference value, and by performing the following actions during an active period in which the technical system is in an active state different from the calibration state:the corrected input conversion unit is used to measure the input quantity, and / or the corrected output conversion unit is used to measure the output quantity.

[0009] The inventive method achieves and combines a multitude of advantageous effects. For example, it makes it possible to correct a potentially incorrectly configured input conversion unit without requiring a complex calibration or adjustment setup. As is well known from the prior art, calibration tests are often carried out in dedicated calibration laboratories. This is no longer necessary in scenarios where the invention can be applied, since calibration can be performed in the setting where the given technical system is already being used. Furthermore, the need to generate and provide reference signals specifically for calibration is eliminated.Particularly with sensors and transducers for measuring high AC voltages, providing realistic reference signals is often associated with considerable difficulties, which are no longer necessary within the scope of the invention. Furthermore, it is possible to perform calibration much more frequently, for example, even during operational downtime, which would not be possible if the sensors to be calibrated had to be transported to a calibration laboratory for this purpose. This increases measurement accuracy and reliability.

[0010] It should be emphasized that the invention makes it possible to calibrate both conversion units. Although calibrating the input conversion unit is a particularly advantageous implementation of the invention, which can be used to address the aforementioned case of a faulty AC voltage measurement in a mains rectifier, it is equally conceivable to correct only the output conversion unit in accordance with the invention, which allows the invention to be applied in a multitude of other applications.

[0011] In practice, it is common for conversion units to have a converter gain to convert a corresponding sensor response into a measured value. In this case, the invention can be implemented particularly easily by increasing or decreasing the converter gain for correction. A converter gain can be understood, in particular, as a multiplicative amplification factor implemented by the conversion unit between the sensor response and the measured value. Depending on the embodiment, the invention preferably allows the converter gain to be increased or decreased by up to 10%, up to 50%, or even up to 100% or more of a current value. Such conversion units can be implemented in the form of analog conversion units, such as operational amplifiers (op-amps), e.g., analog-to-digital converters (ADCs), such as sigma-delta converters, etc., or the conversion units can also be implemented as digital conversion units, for example in the form of a digital scaling element that adapts a sensor response to a given range or to a given format.

[0012] In advantageous embodiments of the invention, an input measurement profile comprising a multitude of input measured values ​​and an output measurement profile comprising a multitude of output measured values ​​can be recorded and compared with each other in the calibration section in order to generate the calibration reference value. This approach allows, on the one hand, the consideration of a larger number of measured values ​​and thus the consideration of more information, and on the other hand, also the application of modern filtering methods for processing the measured values.Specifically, the input measurement profile can be filtered using an input measurement filter before comparison with the output measurement profile, and / or the output measurement profile can be filtered using an output measurement filter before comparison with the input measurement profile, and / or a time course of calibration comparison values ​​can be filtered using a comparison filter before use for correcting the input conversion unit.The input measurement filter and / or the output measurement filter and / or the comparison filter can be implemented in the form of a low-pass filter, a high-pass filter, a band-pass filter, a mean-value filter, a maximum filter, a minimum filter, an absolute value calculation, a scaling factor, or a combination of the aforementioned filters, for example, to suppress measurement noise and / or to eliminate systematic disturbances that arise, for example, from peculiarities and / or dynamic characteristics of the technical system in the calibration state.

[0013] In a further advantageous embodiment of the invention, the calibration reference value can be compared with a predetermined plausibility value to perform a plausibility check, for example, by subtraction or by another comparison method. If the calibration reference value exceeds or falls below the plausibility value, a safety action is triggered, such as an emergency stop, a safe shutdown function, or the issuance of a warning. In this way, it also becomes possible to increase the safety of a technical system in which the method according to the invention is used.

[0014] The method according to the invention can be implemented, in particular, in a calibration device, which preferably comprises suitable hardware capable of performing the method steps according to the invention. Such a calibration device can also, in a particularly preferred manner, be provided as part of a measuring device for determining a measured value.

[0015] As explained earlier, mains rectifiers represent a particularly advantageous application area of ​​the present invention, wherein the mains rectifier can be provided, in particular, as part of a battery emulator or as part of a battery tester. In such scenarios, a calibration state of the mains rectifier can be provided in which the semiconductor switches present in the mains rectifier are open and not switched, and an active state of the mains rectifier can be provided in which the semiconductor switches present in the mains rectifier are switched in order to regulate a DC intermediate circuit voltage output by the mains rectifier to a predetermined intermediate circuit voltage setpoint.In this configuration, a DC link capacitor is preferably provided at the output of the switchable mains rectifier, across which the DC link voltage drops, making it possible to advantageously select a time period as the calibration period in which the DC link capacitor is charged and not loaded.

[0016] The present invention is described below with reference to the Figures 1 to 4 In more detail, the invention is explained, and exemplary, schematic, and non-restrictive embodiments are shown. This includes showing Fig. 1 a state-of-the-art battery emulator with mains rectifier, Fig. 2 a comparison of two measuring chains according to the invention, Fig. 3a-f possible signal waveforms when applying the invention, Fig. 4 a control of a power rectifier using the method according to the invention.

[0017] The present invention is explained below with reference to a battery emulator 1, such as that known from AT 517652 B1 or WO 2013 / 174967 A1. However, it should be noted at this point that the invention can be used not only with battery emulators 1, but in virtually all technical applications that are based on the measurement of two interacting physical quantities and where high accuracy requirements must be met.

[0018] Accordingly, the invention can also be used, for example, in an electric motor, such as a DC motor, a synchronous motor, or an asynchronous motor, in whose phases a phase current is applied as an input variable and which generates a torque as an output variable. The invention can also be used in an internal combustion engine, where an accelerator pedal position is specified as an input variable and a crankshaft torque is generated by the internal combustion engine as an output variable. Alternatively, the invention can be used in test benches of various types, such as a powertrain test bench, where a load torque is specified as an input variable by a load machine ("dynamometer") and a test specimen reacts to the load torque with a test specimen rotational speed as an output variable.However, battery emulators 1 and / or battery testers and, in particular, mains rectifiers 2 for battery emulators 1 and / or battery testers represent particularly suitable applications for the present invention, as will be explained below.

[0019] The in Fig. 1 The battery emulator 1 shown consists of an input-side mains rectifier 2, which is connected to a DC voltage converter 3 via a DC link 9 with a DC link voltage U 0 and a DC link capacitance C 0.

[0020] The mains rectifier 2 in the present case comprises the semiconductor switches T a1, T a2, T b1, T b2, T c1, T c2, which can preferably be implemented as IGBT switches or as SiC MOSFET switches, each with freewheeling diodes arranged in parallel to the semiconductor switches. As will be explained in detail later, it was recognized within the scope of the invention that a switchable mains rectifier 2, such as the one from Fig. 1 , can assume two fundamental states with respect to its transfer behavior, specifically a first state ("calibration state") in which all switches are open and not switched, and a second state ("activity state") in which the semiconductor switches are actively switched.

[0021] The DC-DC converter 3 can be, in particular, a boost or buck converter, which, starting from the DC intermediate circuit voltage U0, allows a different, possibly slowly changing, DC voltage to be set at the output in order to simulate the actual behavior of a battery on a test bench, e.g., to supply an electric motor or an inverter designated as load 4. The battery emulator 1 is supplied with a mains-side AC voltage from an AC mains supply. In the illustrated case of a typical 3-phase network with phases L1, L2, and L3, the mains-side AC voltage uab drops between phases L1 and L2, the mains-side AC voltage ubc drops between phases L2 and L3, and the mains-side AC voltage uac drops between phases L1 and L3. These relationships are well known to a specialist in the field of power electronics.

[0022] A load 4 is provided at the output of the DC-DC converter 3, e.g., an inverter to be tested on a test bench, which is supplied with an output voltage u A provided by the DC-DC converter 3. The DC-DC converter 3 can also be implemented as a multi-phase converter, and in particular as an actively switchable synchronous converter with a number of half-bridges (one half-bridge per phase) using semiconductor switches.

[0023] The battery emulator 1 further includes a control unit 5, which controls the mains rectifier 2, i.e., the switches of the mains rectifier 2, and the DC-DC converter 3, i.e., the switches of the DC-DC converter 3, to generate a desired DC intermediate circuit voltage U 0 and / or a desired output voltage u Asoll, which may be requested, for example, by a higher-level control unit. The control unit 5 can be implemented, for example, as microprocessor-based hardware, as a microcontroller, or as an integrated circuit (ASIC, FPGA).

[0024] Typically, pulse-width modulation (PWM) is used to control the semiconductor switches, which is also well known. The pulse-width modulation (PWM) can be implemented directly in the control unit 5, as shown in Fig. 1Assuming the following, the control unit 5 generates switching signals xa, xb, xc for the semiconductor switches T a1, T a2, T b1, T b2, T c1, T c2 of the rectifier 2 and for the semiconductor switches of the DC-DC converter 3 from the setpoint values ​​of the control system, here for the DC intermediate circuit voltage U 0,setpoint and for the output voltage u Assetpoint. However, PWM can also be implemented in the rectifier 2 itself or in the DC-DC converter 3 itself, so that a control unit 5 would only have to output corresponding control signals, e.g. a desired duty cycle, and only in the rectifier 2 or only in the DC-DC converter 3 would trigger pulses for the respective semiconductor switches be determined from a control signal.

[0025] For the practical implementation of a control system, measurement data must be provided to the controller in a known manner, conveying information about the system to be controlled. To control the present mains rectifier 2, measured values ​​of the two mains-side AC voltages uab and ubc and measured values ​​of the DC intermediate circuit voltage U0 are typically required (in the case of the one in Fig. 1 Assuming a symmetrical three-phase AC network, sometimes only one AC voltage is sufficient; typically, two AC voltages are measured and the third AC voltage is calculated from the two measured AC voltages.

[0026] In many cases, PI controllers, PID controllers, sliding-mode controllers, backstepping controllers, or model predictive control concepts are used to control a battery emulator 1 or a mains rectifier 2. These concepts are implemented in a control unit 5, and in addition to the aforementioned measured values ​​of u ab , u bc , U 0, mathematical models are also provided to the control unit 5, in particular models of the mains rectifier 2 and / or the rectifier 3 and / or the load 4, etc.

[0027] To provide the measured values ​​of the mains-side AC voltages uab, ubc and the DC intermediate circuit voltage U0 required for the control system, the following are required in the Fig. 1In the illustrated case, several components are provided. An input sensor S1 is used to measure the AC voltage uab on the mains supply. The input sensor S1 generates an input sensor response SA1 as a function of the AC voltage uab. This input sensor response SA1 is converted by a first input conversion unit W1 into an input measurement y1 and thus supplied to the control unit 5. Furthermore, an output sensor S2 is provided, which measures the DC link voltage U0 (corresponding to an output u2 of the mains rectifier 2) and generates an output sensor response SA2 as a function of the DC link voltage U0. An output conversion unit W2 converts the output sensor response SA2 into an output measurement y2, which is also supplied to the control unit 5 and can be used there for control purposes.The input sensor S1 and the input conversion unit W1 form a first measuring chain MK1, which in this case functions as a voltage measuring chain ("voltmeter"). The output sensor S2 and the output conversion unit W2 accordingly form a second measuring chain MK2, which also functions as a voltage measuring chain, i.e., a voltmeter. The same applies to the third chain. Fig. 1 The illustrated measuring chain MK 3 comprises a third sensor S3 and a third conversion unit, which measures the additional mains-side AC voltage u bc. Such a third measuring chain MK 3 is advantageous in the practical implementation of a control system for a mains rectifier 2 or the invention, but is not strictly necessary.

[0028] As explained at the beginning, the control of mains rectifiers 2 often presents the problem that mains-side AC voltages uab, ubc are not measured with sufficient precision. Measurement inaccuracies in AC-side voltage measurements can significantly impair the control quality of a mains rectifier 2. For example, incorrect information about mains-side AC voltages can lead to incorrect actuation times of the semiconductor switches Ta1, Ta2, Tb1, Tb2, Tc1, Tc2, which may remain open for too long because a controller implemented in the control unit 5 infers excessively low currents based on insufficient measured values. In severe cases, such scenarios can even lead to damage to the components shown (intermediate link capacitor C0, load 4, ...).

[0029] Among other things, these problems are solved, or at least mitigated, by the present invention. To this end, the invention provides for a comparison of the aforementioned measured values ​​y1 and y2. Since the present battery simulator 1 is indeed a preferred application, but the invention is by no means limited to battery simulators 1 and / or mains rectifiers 2, the inventive procedure is described using the following examples: Fig. 2 First, a general description is given. For this purpose, instead of the mains rectifier 2 considered above, a general technical system 2 is assumed, via which an input quantity u 1 to be measured (e.g. a mains-side AC voltage u ab ) is related to an output quantity u 2 to be measured (e.g. a DC intermediate circuit voltage U 0 ).

[0030] As demonstrated by Fig. 2As can be seen, an input measurement y1 and an output measurement y2, each acquired by means of a measurement chain MK1, MK2 described above, consisting of sensor S1, S2 and transducer W1, W2, are compared with each other, preferably subtracted from each other, thereby generating a calibration reference value ey. It should be noted that the invention is not limited to the use of subtraction for comparison. For example, to implement the invention, it could also be simply monitored whether one measurement is greater than another, or whether the measurements differ by more than a predetermined threshold. The input transducer W1 is then corrected to a corrected input transducer W1 based on the calibration reference value ey, and subsequently, the corrected input transducer W1 is used to measure the input quantity u1.For the method according to the invention, it is crucial that the acquisition of at least one input measurement y1, the acquisition of at least one output measurement y2 for comparison, and the comparison itself take place in a temporal calibration phase in which the technical system 2 is in a calibration state, and that in a temporal active phase following the temporal calibration phase, in which the technical system 2 is in an active state different from the calibration state, measurements are taken using the corrected input conversion unit W1, without any further modification or correction of the input conversion unit W1. The correction of the conversion units preferably takes place in the calibration phase, but in principle, correction in the active phase is also conceivable.A calibration device in which the inventive method for calibrating an input conversion unit is implemented, or a measuring device comprising such a calibration device, may preferably be designed to recognize a calibration state or an active state itself, for example by suitable analysis of the recorded measured values, or it may have an interface so that the presence of a corresponding state can be communicated to the calibration device, for example by an operator or by a suitable control unit.

[0031] The procedure described above is based on the understanding that many technical systems require precise measurement of their input and output variables and can assume at least two (system) states. One state may be ideally suited for calibration, while the other is inaccessible. This is particularly true when the input variable u1 is related to the output variable u2 according to a transfer ratio determined by the state of the technical system 2, and the transfer ratio in the calibration section corresponds to a predetermined, constant ratio, while in the active section it corresponds to a variable, and especially an unpredictably variable, transfer ratio.The calibration state that the system assumes during the calibration phase is therefore characterized by a predictable, deterministic transfer characteristic that remains constant or at least largely constant. Under ideal conditions, "constant" would mean that the transfer characteristic does not change at all. However, under real, practical conditions, minor fluctuations are always present, for example, minor fluctuations in the characteristics of components installed in the technical system 2, such as electrical resistors, other electrical components, or mechanical components. Therefore, within the scope of these considerations, a transfer characteristic of a technical system, such as the gain of a technical system, is considered constant if it changes during operation only to the extent caused by minor component fluctuations.This means that the amplification fluctuates by less than 1%, less than 0.1%, or less than 0.01%, for example. In contrast, in an active state, unpredictable changes of more than 25%, more than 50%, or more than 75% of the amplification can occur.

[0032] The inventive method achieves and combines a multitude of advantageous effects. For example, it makes it possible to correct a potentially incorrectly configured input conversion unit W1 without requiring a complex calibration or adjustment setup. As is well known from the prior art, calibration tests are often carried out in dedicated calibration laboratories. This is no longer necessary in scenarios where the invention can be applied, since calibration can be performed in the setting where the given technical system is already being used. Furthermore, the need to generate and provide reference signals specifically for calibration is eliminated.Particularly with sensors and transducers for measuring high AC voltages, providing realistic reference signals is often associated with considerable difficulties, which are no longer necessary within the scope of the invention. Furthermore, it is possible to perform calibration much more frequently, for example, even during operational downtime, which would not be possible if the sensors to be calibrated had to be transported to a calibration laboratory for this purpose. This increases measurement accuracy and reliability.

[0033] If the input conversion unit W1 has a converter gain to convert the input sensor response SA1 into the input measured value y1, increasing or decreasing the converter gain to correct the input conversion unit W1 often proves to be an efficient implementation of the invention. To determine the increase or decrease, a scaling factor can be specified, for example, which maps the calibration reference value ey to a magnification value for increasing or decreasing the input conversion unit W1. However, in principle, in the case of alternatively constructed or more complex conversion units, it is also possible to implement a correction according to the invention at other points. For example, it is also conceivable to adjust the clock frequency of a digitally implemented conversion unit or to change the dead time of a conversion unit. The invention offers considerable flexibility in this respect.

[0034] Furthermore, the inventive method can easily be supplemented and advantageously extended in several places. For example, it is possible to compare the calibration reference value ey with a predetermined plausibility value to perform a plausibility check and, if the calibration reference value ey exceeds or falls below the plausibility value, to trigger a safety action. A safety action could, for example, be the output of a warning signal informing a user that one of the existing sensors is defective, or the technical system could be deactivated or switched to a safety mode.

[0035] As also based on Fig. 2As can be seen, all signals occurring within the framework of the method according to the invention can be filtered. This is sometimes necessary to make different measurement signals comparable, e.g., DC voltages with AC voltages, for which purpose a peak value of an AC voltage can be determined and compared with a DC voltage by means of suitable filtering. Specifically, for the purpose of filtering in the calibration section, an input measurement curve yt1 comprising a plurality of input measurement values ​​y 1,k , y 1,k+1 , y 1,k+2 , ... and an output measurement curve yt2 comprising a plurality of output measurement values ​​y 2,k , y 2,k+1 , y 2,k+2 , ... can be recorded (the index k, as is customary in digital signal processing, represents a time index to describe integer multiples of a sampling time T d , i.e., tk = k*T d ) and the entire curves or at least parts of the curves can be compared with each other, i.e.,A large number of input and output measurements are compared to generate the calibration reference value ey. To filter such time profiles, an input measurement filter F1 and / or an output measurement filter F2 and / or a reference measurement filter F3 can be provided to filter the calibration reference value ey, preferably in the form of a low-pass filter, a high-pass filter, a band-pass filter, a mean value filter, a maximum filter, a minimum filter, an absolute value calculation, a scaling factor, or a combination of the aforementioned filters.In the previously mentioned case of comparing an AC voltage with a DC voltage, an input measurement filter F1 can, for example, also be a non-linear filter that first establishes an absolute value and then determines a maximum peak value for each period, which is compared with a DC voltage measured at the output. A comparison measurement filter F3 can, in particular, contain a suitable scaling factor to convert the calibration reference value ey to a suitable signal for calibrating the conversion units.

[0036] Possible time courses of an input measurement y1 and an output measurement y2 are shown in Fig. 3 shown. Figures 3aFigures 3b (uncalibrated) and 3b (calibrated) show two time profiles of an input measurement y1 and an output measurement y2, which are constant in their average. Before the calibration according to the invention, they are shown side by side (i.e., shifted along the vertical y-axis), and after the calibration according to the invention, they are shown one above the other. Figures 3c (Uncalibrated) and 3D (calibrated) further illustrate the practically important case of an input AC quantity, e.g., an AC voltage with an offset, and an output DC quantity, which can, for example, simply correspond to the offset. Calibration ensures that the output DC quantity and the offset of the input quantity match. Figures 3e(uncalibrated) and 3f (calibrated) further illustrate the case, particularly important for an application of the invention with a rectifier 2, of comparing peak values ​​y1 with a DC quantity y2. By reducing a measurement gain, in the case of the Figures 3e and 3f The peak values ​​y1 are aligned with the constant quantity y2.

[0037] As mentioned, it is possible to use the invention in a particularly advantageous way for operating a mains rectifier 2. All of the above considerations can therefore be applied to a mains rectifier 2 as already described in Fig. 1 as shown below, they can be transferred, as illustrated below. Fig. 4 will be explained.

[0038] Specifically, it shows Fig. 4This is a particularly advantageous application of the method according to the invention, wherein a switchable mains rectifier 2 comprising switchable semiconductor switches T a1, T a2, ... is provided as the technical system 2. A calibration state of the mains rectifier 2 is provided in which the semiconductor switches T a1, T a2, ... in the mains rectifier 2 are open and not switched. A time interval in which the intermediate circuit capacitor C 0 is charged and, in particular, not yet loaded, can be selected as the calibration period, especially if it is particularly preferred. An active state of the mains rectifier 2 can be provided in which the semiconductor switches T a1, T a2, ... in the mains rectifier 2 are switched in order to regulate a DC intermediate circuit voltage U 0 output by the mains rectifier 2 to a predetermined intermediate circuit voltage setpoint.In the active state, the intermediate circuit capacitor C 0 is therefore loaded, meaning in particular that electrical charge is taken from the intermediate circuit capacitor C 0 towards the load 4 and new electrical charge is supplied, in contrast to the calibration section, in which preferably no load takes place.

[0039] Calibration sections arise in the case of the in Fig. 4The power supply rectifier 2 shown is, in particular, connected to the AC network after AC-side contactors (not shown) have been connected to and disconnected from the AC network to establish a connection between the power supply rectifier 2 and the AC network, but the semiconductor switches T a1, T a2, ... in the power supply rectifier 2 are not yet actively switched, so that only a passive charging of the DC link capacitor C 0 occurs via the freewheeling diodes in the power supply rectifier 2 (the power supply rectifier 2 is passive), the passive charging of the DC link capacitor C 0 is already complete, but the DC link is still unloaded. To check whether the DC link capacitor C 0 is charged, it can be checked whether a corresponding charging process has been completed. For this purpose, it can be monitored, for example, whether the DC link voltage U 0 across the DC link capacitor C 0 remains constant, i.e.,In practice, the DC link voltage U0 across the DC link capacitor C0 must fluctuate by less than a predefined threshold, e.g., by less than 10V, 5V, 1V, or 0.1V. Monitoring the current flows in each phase of the rectifier 2 could also be considered (ideally, these currents disappear if the DC link voltage U0 is constant and the DC link capacitor C0 is charged, or at least become very small in practice). Fluctuations in the DC link voltage U0, or short-term high values ​​that have not yet dropped back to the passive voltage level due to previous active operation of the rectifier 2, can thus be detected. Consequently, the voltage stability of the DC link voltage U0 is monitored.Any problems that arise from this can be countered by waiting and restarting the measurement / calibration (automatically).

[0040] As in Fig. 4 As shown, two subunits 51, 52 can be provided in the control unit 5 to implement the invention. In the first subunit 51, as indicated, the switching signals xa, xb, xc for switching the semiconductor switches T a1, T a2, ... can be determined, in particular by means of a controller implemented in the first subunit 51. In the second subunit 52, the correction of the input conversion unit W 1 according to the invention can be carried out. As shown in Fig. 4As shown, the second subunit 52 can determine the correction signals z1, z2, z3 for this purpose. These correction signals z1, z2, z3 can be multiplicative correction signals z1, z2, z3, which in particular scale a converter gain, or they can be additive correction signals z1, z2, z3, which are added to a converter gain, for example, or they can even determine a completely new converter gain that replaces an old one. As also shown in Fig. 4 As indicated, the aforementioned subunits 51 and 52 can of course communicate with each other, so that, for example, the second subunit 52 can inform the first subunit 51 that a calibration process is currently taking place and that active switching should not yet begin.

[0041] To determine a new converter gain, the formula 2*U0 / USS can be used, where USS / 2 is half the peak-to-peak value of the AC voltage at its fundamental frequency component (Europe: 50 Hz, USA: 60 Hz). If USS / 2 is greater than U0, the converter gain is less than 1.0, which can then be used to scale down the line-side voltages accordingly when measuring them. Conversely, if USS / 2 is greater than U0, the converter gain would be greater than 1.0. The freewheeling diodes present in the passive rectifier are advantageously and simply neglected in this calculation, as the resulting error is also negligible. If these diodes were included, they would reduce the passive rectified value by two diode forward voltages.

[0042] Applied to a Fig. 4In this situation, the idea of ​​the invention can therefore be summarized as follows. The DC intermediate circuit voltage after passive rectification via the diodes is used to provide a reference value for the peak value of the AC voltage. Both the passive DC rectification voltage and the peak value of both line-to-line voltages are averaged, then an error is calculated, and finally the measurement gain is corrected for this error.

Claims

1. Method for determining a measured value (y1, y2), comprising: - an input sensor (S1) for measuring an input quantity (u1) of a technical system (2) and for outputting an input sensor response (SA1) as a function of the input quantity (u1), - an input conversion unit (W1) for converting the input sensor response (SA1) into an input measured value (y1), - an output sensor (S2) for measuring an output quantity (u2) of the technical system (2) and for outputting an output sensor response (SA2) as a function of the output quantity (u2), - an output conversion unit (W2) for converting the output sensor response (SA2) into an output measured value (y2), characterized by the fact that In a time-based calibration period, during which the technical system (2) is in a calibration state, an input measurement (y1) and an output measurement (y2) are recorded and compared to obtain a calibration reference value (e) yto generate that the input conversion unit (W1) depending on the calibration reference value (e y ) is corrected to a corrected input conversion unit (W1) and / or the output conversion unit (W2) depending on the calibration reference value (e y ) is corrected to a corrected output conversion unit (W2), and that in a temporal active section in which the technical system (2) is in an active state different from the calibration state, the corrected input conversion unit (W1) is used to measure the input quantity (u1), and / or the corrected output conversion unit (W2) is used to measure the output quantity (u2).

2. Method according to claim 1, characterized by the fact thatthe input conversion unit (W1) has a converter gain to convert the input sensor response (SA1) into the input measurement value (y1) and / or the output conversion unit (W2) has a converter gain to convert the output sensor response (SA2) into an output measurement value (y2), and that the converter gain for correction depending on the calibration reference value (e y ) is enlarged or reduced.

3. Method according to any one of the preceding claims, characterized by the fact that the input variable (u1) is related to the output variable (u2) according to a transfer ratio of the technical system (2) determined by a state of the technical system (2), wherein the transfer ratio in the calibration section corresponds to a predetermined, constant transfer ratio and in the active section corresponds to a variable transfer ratio.

4. Method according to any one of the preceding claims, characterized by the fact that In the calibration section, an input measurement sequence (yt1) comprises a multitude of input measured values ​​(y 1,k , y 1,k+1 , y 1,k+2 , ...) and an output measurement profile (yt2) comprising a multitude of output measurements (y 2,k , y 2,k+1 , y 2,k+2 , ...) are recorded and compared with each other to determine the calibration reference value (e y to generate.

5. Method according to claim 4, characterized by the fact that The input measurement curve (yt1) is filtered by means of an input measurement filter (F1) before comparison with the output measurement curve (yt2). and / or that The output measurement curve (yt2) is filtered by means of an output measurement filter (F2) before comparison with the input measurement curve (yt2). and / or that a time course of calibration reference values ​​(e y ) is filtered before being used to correct the input conversion unit (W1) using a comparison filter (F3).

6. Method according to claim 5, characterized by the fact that the input measurement filter (F1) and / or the output measurement filter (F2) and / or the comparison filter (F3) is implemented in the form of a low-pass filter or in the form of a high-pass filter or in the form of a band-pass filter or in the form of an average value filter or in the form of a maximum filter or in the form of a minimum filter or in the form of an absolute value calculation or in the form of a scaling factor or in the form of a combination of the aforementioned filters.

7. Method according to any one of the preceding claims, characterized by the fact that the calibration reference value (e y ) to perform a plausibility check against a predetermined plausibility value and that if the calibration reference value (e y ) if the plausibility value exceeds or falls below the threshold, a security action is triggered.

8. Method according to any one of the preceding claims, characterized by the fact thatA switchable mains rectifier (2) is provided as a technical system (2) with a mains-side AC voltage as input variable (u1) and a DC intermediate circuit voltage as output variable (u2) or that an electric motor is provided as a technical system (2) with an AC phase current as input variable (u1) and a torque generated by the electric motor as output variable (u2) or that an internal combustion engine is provided as a technical system (2) with an accelerator pedal position as input variable (u1) and a crankshaft torque as output variable (u2) or that a powertrain test bench is provided as a technical system (2) with a load torque as input variable (u1) and a test specimen speed as output variable (u2).

9. Method according to claim 8, characterized by the fact that comprising a switchable mains rectifier (2) and switchable semiconductor switches (T) a1 , T a2, ...) is provided as a technical system (2), wherein a state of the mains rectifier (2) is provided as the calibration state in which the semiconductor switches present in the mains rectifier (2) (T a1 , T a2 , ...) are open and are not switched and wherein an active state of the mains rectifier (2) is provided in which the semiconductor switches present in the mains rectifier (2) are switched in order to regulate a DC intermediate circuit voltage output by the mains rectifier (2) to a predetermined intermediate circuit voltage setpoint.

10. Method according to claim 9, characterized by the fact that A DC link capacitor (C0) is provided at the output of the switchable mains rectifier (2), across which the DC link voltage drops. and that The calibration period is a time interval in which the intermediate circuit capacitor (C0) is charged and preferably not loaded.

11. Method according to any one of claims 8 to 10, characterized by the fact that A switchable mains rectifier (2) is provided as a technical system (2), and the switchable mains rectifier (2) is provided as part of a battery emulator (1) or as part of a battery tester, wherein the mains rectifier (2) is connected on an input side of the mains rectifier (2) to an AC supply network that provides a mains-side AC voltage for the mains rectifier (2), and is connected on an output side of the mains rectifier (2) to an intermediate circuit capacitor (C0) for stabilizing a DC intermediate circuit voltage (U0) output by the mains rectifier (2), wherein the intermediate circuit capacitor (C0) is further connected to a DC voltage converter (3) to convert the DC intermediate circuit voltage (U0) into an output voltage (u A to convert.

12. Method according to claim 11, characterized by the fact thata PI controller or a PID controller or a sliding-mode controller or backstepping controller or a model-based control with a model of the grid rectifier (2) is used to control the grid rectifier (2), wherein measured values ​​of the grid-side AC voltage are used as input measured value (y1) and measured values ​​of the DC intermediate circuit voltage (U0) are used as output measured value (y2) in the control.

13. Calibration device for calibrating an input conversion unit (W1) and / or an output conversion unit (W2), wherein the calibration device is configured to compare an input measurement value (y1) and an output measurement value (y2) in order to determine a calibration reference value (e y ) to generate and the input conversion unit (W1) depending on the calibration reference value (e y) to correct to a corrected input conversion unit (W1) and / or the output conversion unit (W2) depending on the calibration reference value (e y) to a corrected output conversion unit (W2), wherein - the input measured value (y1) can be generated by an input sensor (S1) measuring an input quantity (u1) of a technical system (2) which is in a calibration state and outputting an input sensor response (SA1) as a function of the input quantity (u1) and the input conversion unit (W1) converting the input sensor response (SA1) into the input measured value (y1), - the output measured value (y2) can be generated by an output sensor (S2) measuring an output quantity (u2) of the technical system (2) which is in a calibration state and outputting an output sensor response (SA2) as a function of the output quantity (u2) and the output conversion unit (W2) converting the output sensor response (SA2) into the output measured value (y2).

14. Measuring device for determining a measured value (y1, y2), comprising: - an input sensor (S1) for measuring an input quantity (u1) of a technical system (2) and for outputting an input sensor response (SA1) as a function of the input quantity (u1), - an input conversion unit (W1) for converting the input sensor response (SA1) into an input measured value (y1), - an output sensor (S2) for measuring an output quantity (u2) of the technical system (2) and for outputting an output sensor response (SA2) as a function of the output quantity (u2), - an output conversion unit (W2) for converting the output sensor response (SA2) into an output measured value (y2), - a calibration device according to claim 13, which is configured to provide an input measured value (y1) and an output measured value (y2) that are available in a time-based calibration period during which the technical system (2) is in a calibration state, be detected,to read in and compare them to obtain a calibration reference value (e, y ) to generate, and the input conversion unit (W1) depending on the calibration reference value (e y ) to correct to a corrected input conversion unit (W1) and / or the output conversion unit (W2) depending on the calibration reference value (e y ) to correct to a corrected output conversion unit (W2) in order to measure the input quantity (u1) using the corrected input conversion unit (W1) and / or the output quantity (u2) using the corrected output conversion unit (W1) in a time active section in which the technical system (2) is in an active state different from the calibration state.

15. Measuring device according to claim 14, wherein a mains rectifier (2) is provided as a technical system (2).

Images